BACTERIOME ASSOCIATED WITH Rhizophora mangle SEDIMENTS WITHIN BRAZIL SEMI-ARID MANGROVES

Microorganisms play important roles in nutrient cycling in mangrove ecosystems and knowledge on the plant/microorganism association is essential to better understand the functioning of this environment. Rhizophora mangle is the dominant tree species within Brazilian mangroves and little information is available on the microbiota associated with this plant species. In this context, the aim of this study was to survey the taxonomic diversity of bacteria in the R. mangle root zones in mangroves within the semi-arid region of Northeast Brazil submitted to different human disturbances, intending to determine the bacterial fingerprint associated with this habitat. The total DNA extracted from sediments of different mangroves was pooled and used for construction of 16S rDNA cloning libraries, which resulted in 663 sequences with an average size of 809 bp. All mangroves were rich in different phyla of the Bacteria domain, with Acidobacteria, Bacteroidetes, Chloroflexi, and Proteobacteria being detected in all locations. Proteobacteria was dominant in all mangroves, and it was mainly represented by Alpha, Delta, and Gammaproteobacteria. The greatest richness was found in the Timonha river mangrove, with 13 phyla, a location considered more preserved compared to other mangroves. The lowest richness was found in Ceará river mangrove, with only seven phyla. This mangrove is threatened by intense urbanization. The results clearly showed that the taxonomic diversity of bacteria from mangroves subjected to intense urbanization have decreased, highlighting the risks of these changes for the functioning of important microbe-mediated processes and related ecosystem services. foi encontrada no manguezal do rio Timonha, com 13 filos, local considerado mais preservado em comparação com outros manguezais. A menor riqueza foi encontrada no manguezal do rio Ceará, com apenas sete filos. Esse manguezal é ameaçado por intensa urbanização. Os resultados mos­ traram claramente que a diversidade taxonômica de bactérias de manguezais submetida à intensa urbanização diminuiu, evidenciando os riscos dessas alterações para o funcionamento de impor­ tantes processos mediados por micro­organismos e serviços ecossistêmicos relacionados. Palavras-chave: manguezais brasileiros, semiárido, Rhizophora rRNA 16S, biblioteca clones. mangroves . The results endorsed a general pattern, a core of bacteria phylum-groups in mangroves, regardless of the local environment characteristics, and high redundancy of bacterial genera involved in the sulfur metabolism, especially sulfate-reducing bacteria. These organisms proliferate in habitats rich in partially oxidized substrates and anaerobic conditions, typically found inside red mangrove habitats. Furthermore, our study presents the identification of a signature of bacterial genera inside red mangrove habitats in the Northeastern Brazilian mangroves.


INTRODUCTION
Vegetation that is exclusively found in mangroves comprises about 73 species within 28 genera, which present a number of different adaptations allowing them to survive in this salty, muddy and anoxic coastal environment (Duke et al., 2006;Lo et al., 2014). In Brazil, this exclusive vegetation is represented by the genera Avicennia, Laguncularia and Rhizophora, with the latter prevailing throughout the coastal zone (Spalding;Kainuma & Collins, 2010).
Rhizophora is the predominant and the most widespread genus of the family Rhizophoraceae (Lo et al., 2014). Although globally distributed, only six species and approximately an equal number of hybrids have been described for this genus: R. apiculata, R. mucronata and R. stylosa in Indo-West Pacific, and R. mangle, R. racemosa and R. samoensis in Atlantic-East Pacific mangroves (Lo et al., 2014).
R. mangle, also known as red mangrove, is the dominant species in tropical coastal areas of the Atlantic Ocean, Pacific America and in several tropical islands at the South-Western Pacific Ocean (Duke et al., 2006;Lo et al., 2014;Spalding;Kainuma & Collins, 2010). The red mangrove has an intricate prop root system that supports its fixation in muddy soils, and its respiration and aeration during periods of complete submersion (Duke et al., 2006).
Since the red mangrove is predominant in tropical mangroves, it is assumed that this plant might have an essential role for the ecosystem, and it has been hypothesized that microbial community inhabiting this environment is very specific due to selection over time (Cleary et al., 2012;Dias et al., 2012). Despite that intimate association, plant effects on microbial diversity and vice-versa is little understood in mangroves (Gomes et al., 2010). In terrestrial plants, such interaction is recognized as the rhizospheric effect; i.e., the effect caused by plants providing nutrient-rich exudates (Höflich;Wiehe & Kühn, 1994) and therefore select for certain taxa and/or functional groups of microorganisms (Kumar et al., 2007). On the other side, microorganisms are essential for plant growth, protection against pathogen and organic matter turnover (Narula;Kothe & Behl, 2009).
Soil bacteriomes associated to R. mangle of the semi-arid coast of Brazil were reported as rich and complex repositories of diversity and complexity when compared with R. mangle forests in the North or South of Brazil (Tavares et al., 2021). The semi-arid coast displays important features that can explain such findings, such as water deficit caused by low rainfall, water impoundment by dams, and elevate evapotranspiration. Those features contribute to saline intrusion and landward migration of such mangrove forests, a temporal trend recently reported by Godoy and Lacerda (2015). Considering that the composition of sediment microbial communities can also suffer the influence of plant diversity and genotype related to rhizodeposits types and amounts, root architecture, and plant physiological activity (Craig et al., 2020;Philippot et al., 2013), the analysis of the R. mangle associated microbiota can help to understand how variable those communities are in a spatial scale and to discern among abiotic drivers of variation.
This study aimed to conduct the first survey of the benthic bacterial community under Rhizophora mangle forests across eight mangroves within the semi-arid coast of Northeast Brazil, intending to determine the bacterial fingerprint associated with that habitat. This knowledge traces a baseline regarding the bacterial diversity associated with R. mangle, which can help future studies associated with ecosystem monitoring of the effects of anthropogenic actions, such as the impacts of oil spills, which recently reached the coast of Brazil (Soares et al., 2020). Knowing those communities is strategic to use microbial diversity with creativity in a science-based manner to deal with Anthropocene's problems in our changing biosphere.

MATERIALS AND METHODS
Sediment samples were taken from eight mangroves across the 573 km of Ceará state's coastal zone, on the semi-arid coast of Brazil (Fernandez et al., 2019). The study comprised samples from Icapuí (ICA), Malcozinhado (MAL), Cocó (COC), Ceará (CEA), Aracatiaçu (ARA), Acaraú (ACA), Coreaú (COR) and Timonha (TIM) mangroves. CEA, COC, and MAL are located in the metropolitan region of Fortaleza, the biggest city of the State of Ceará, and are intensively subjected to urbanization and sewage disposal. ARA, ACA, and COR are located on the west coast of the state and are threatened mostly by shrimp farming. TIM was considered as the reference mangrove, since it is located in the environmental protection area of the Parnaiba River, at the extreme west of the State of Ceará. ICA is located on the extreme east coast in a risk area for oil spills due to its proximity to petroleum exploitation reservoirs.
Sediments were collected within the prop roots of R. mangle in the 0-20 cm superficial layer, at low tide (0.0-0.2 m), during the wet season (between March and May 2010). For the molecular analysis, samples were stored in sterile flasks, labeled, and placed in an ice-cooled box. Sediments for physicochemical analyses were packaged in plastic bags and taken to the laboratory. Salinity and pH were measured in the sediment interstitial water. The remaining samples were dried for organic matter and particle size analyses.
Organic matter content (OMC) was analyzed by the loss on ignition protocol (Schulte & Hopkins, 1996). In order to perform the particle size analysis, samples of dry sediments were weighed (100 g) in triplicates and washed in a 0.062 mm mesh sieve to remove the silt-clay from the sand fraction. The sand retained on the sieve was recovered, oven-dried, and re-weighed. The silt-clay content was calculated by subtracting the final and initial weights. PC-ORD (Mccune & Mefford, 2011) was used to generate the cluster analysis of the abiotic factors based on Euclidian distance. The variables were also analyzed by ANOVA with p < 0.001 and Tukey a posteriori test.
The metagenomic DNA was isolated from 0.5 g of sediments using the PowerSoil TM DNA Isolation kit (MoBio, Carlsbad, CA, USA). 20 ng of purified DNA was amplified by PCR using the universal primers 27F and 1525R (Lane, 1991), covering nearly the entire 16S rRNA gene (E. coli 16S rRNA gene sequence). The amplicons for each mangrove were cloned into the pGEM-T easy cloning vector (Promega, Madison, WI, USA) and introduced into Escherichia coli Top10F' (Invitrogen, Carlsbad, CA, USA) by electroporation, according to the protocol provided by the manufacturer. The plasmids containing the 16S rRNA gene were extracted from the clones by the alkaline lysis method (Sambrook & Russel, 2001) and sequenced by Macrogen Inc. (Seoul, Korea) using the universal primer T7 promoter in an ABI 3730 (Applied Biosystems) with ABI PRISM BigDyeTM terminator cycle sequencing kit and AmpliTaq DNA polymerase (Applied Biosystems, Foster City, CA, USA), following the protocols supplied by the manufacturer.
Electropherograms were analyzed, and sequences were trimmed and selected for high-quality sequences at Ribosomal Data Project Pipeline (rdp.cme.msu.edu). Sequences longer than 300 bp and with more than 80% of bases with Phred > 20 were selected to constitute the final dataset. Sequences were assigned to operational taxonomic units (OTU) with Mothur software (version 1.32.0) (Schloss et al., 2009), using a naïve Bayesian classification based on the reconstructed SILVA SEED database (release 111) (Quast et al., 2013) as a reference dataset for classifying sequences from the phylum to genus level. A similarity of 98% between sequences was considered to assemble them into OTU at the species level (Nemergut et al., 2011). Chimera checking was performed using the chimera. uchime algorithm (Edgar et al., 2011) with Mothur and SILVA-Gold as the references. A set of 663 sequences passed the quality filtering steps and was used to calculate the richness estimators and diversity indexes. Mothur was also used to build clusters based on relative abundance at different hierarchical levels.

Physicochemical characteristic of semi-arid mangrove sediments
In the present study, we examined the geographic distribution of benthic bacterial communities associated with red mangrove over 573 km of the coastal zone under the semi-arid climate of Northeast Brazil. The eight studied mangroves were chosen taking into consideration local environmental characteristics and the main anthropic impacts acting on each habitat. The environmental variables (pH and salinity), sediment characteristics (organic matter and silt+clay contents), and GPS coordinates of the eight sampling sites are shown in Table I. The variables were analyzed by ANOVA with p < 0.001 and Tukey a posteriori test, which exhibited a statistically significant difference of the silt+clay and organic matter contents between COR and the other mangroves. Also, silt+clay content was shown to increase from the east (ICA) to the west (TIM), probably due to the regional drift current that flows predominantly from the east to the west. Organic matter was mostly related to silt+clay contents, and this fact was clearly seen in COR, where both were higher. Figure 1 shows the clustering based on Euclidean distance for the abiotic variables, where COR is separated from the other mangroves. Also, ACA and TIM clustered together, with 73% of similarity, separating themselves from the group formed by the other mangroves.

Phylum-level bacterial taxonomic composition
After quality checking, 663 sequences (Phred > 20) of 16S rRNA partial gene were obtained, reaching an average of 83 sequences per sample, with a mean size of 809 bp (NCBI PopSet 846337277) (Table II). Richness (Chao1 estimator) and diversity (Shannon index) were highest in TIM, a pristine mangrove, and lowest in the most urbanized mangrove (CEA). According to the observed OTUs and Chao1 richness estimator, there are significant differences between the more preserved and urban mangroves (p < 0.05). The ecological indexes showed that those more altered urban mangroves presented lower richness and diversity when compared to the preserved mangrove, signaling an urgent need for conservation actions.
The retrieved sequences were affiliated to at least 19 phyla with remarkable dominance (> 1% relative abundance): Proteobacteria, Acidobacteria, Actinobacteria, Bacteroidetes, Chloroflexi, Planctomycetes, Deferribacteres, and Firmicutes in all mangroves. Members of these eight phyla make up an average of 94.75% of the libraries. In addition, members of other phylum-level lineages with relative abundance lower than 1%, such as Chlamydiae, Chlorobi, Cyanobacteria, Lentisphaerae, Nitrospira, Gemmatimonadetes, Spirochaetes, Synergistetes, Tenericutes, Verrucomicrobia, and WS3, were detected in the dataset. Only 6.38% of total sequences could not be affiliated to any known bacterial group, evidencing that the study dataset possesses sequences with high quality and length, which provided good coverage of the 16S rRNA gene enabling the classification of a great part of the sequences.
TIM and COR exhibited 13 phyla, followed by ICA (12 phyla), COC and MAL (10 phyla each), and CEA (seven phyla). The preserved mangrove (TIM) shared 10 phyla with its neighbor COR (threatened by intense shrimp farming), and both shared nine phyla with ICA, located in the extreme of the east coast. The adjacent mangroves ARA and ACA, which are quite threatened by shrimp farming, shared seven out of the total of nine phyla found in each. Proteobacteria, Acidobacteria, Bacteroidetes, and Chloroflexi were detected in all locations, comprising at least 71.6% of total phyla relative abundance. Proteobacteria alone ranged from 40 up to 78% in ICA and CEA, respectively. The dominance of Proteobacteria in mangrove sediments is remarkable, regardless of the methodology applied in the study or of the pressures to which the environment is subjected (Figure 2). This dominance was reinforced by Illumina 16S rRNA gene amplicon sequencing, which pointed to the dominance of Proteobacteria (mainly by Deltaproteobacteria and Gammaproteobacteria, which are part of the core microbiome of mangroves worldwide) in the soil associated with R. mangle in the North, Northeast, and South of Brazil (Tavares et al., 2021). Although the high throughput sequencing used by those authors revealed the presence of other phyla, such as Euryarchaeota and Epsilonbacteraeota, in mangroves from the North and South of Brazil as well as in Icapui mangrove (ICA in the present study), the dominance of Proteobacteria is unquestionable. Andreote and colleagues (2012) obtained similar results at phylum-level analyzing sediments from different areas of mangroves in the Southeast region of Brazil (Cananeia and Bertioga, SP), where the relative abundance of Proteobacteria ranged from 47.1 to 56.3%. Pureza et al. (2013) also reported a similar phylum-core in a mangrove from Amazonia (Bragança, PA), and Ghosh et al. (2010) showed that the relative abundance of Proteobacteria was greater than 50% in an Indian mangrove (Ghosh et al., 2010;Pureza et al., 2013). Fierer et al. (2012) has pointed out that Proteobacteria are ubiquitous in soils, and only their relative abundances and their specific representatives tend to vary between different ecosystems (Fierer et al., 2012).
Members of Acidobacteria are phylogeneic and physiologically diverse and ubiquitous in soils, with few described strains (Jones et al., 2009;Wang et al., 2012). In this study, Acidobacteria ranged from 3.8 to 12% (CEA and ARA, respectively), which contrast with the low abundance previously reported in mangroves from Southeastern Brazil (Andreote et al., 2012). Studies suggest that the abundance of this phylum is significantly BACTERIOME ASSOCIATED WITH Rhizophora mangle SEDIMENTS WITHIN BRAZIL SEMI-ARID MANGROVES increased when pH is lower than 5.5, although in this study the pH variation has ranged around 7.0. Wang and colleagues (2012) reported the abundance of Acidobacteria ranging between 4 up to 6% in no acidic freshwater samples and suggested that not only pH drives the abundance of this group. Previous studies reported a higher abundance of Acidobacteria (29% and 54%) in the Brazilian Atlantic forest (Bruce et al., 2010), and in pasture soils (7% and 14%) (Rocha et al., 2013). Therefore, R. mangle from the semi-arid region of Brazil emerges as a new habitat for Acidobacteria prospection. Within the classes of Acidobacteria, Gp10, Gp17, Gp21, and Gp23 were the most representative in our dataset, although other subdivisions were also found. Studies point out that subdivisions Gp1, Gp3, Gp4, Gp6, and Gp18 are enriched in freshwater environments, while subdivisions Gp10, Gp21, Gp22, and Gp26 are more abundant in marine sediments (Orcutt et al., 2011;Zhang et al., 2015). When analyzing soils and sediments collected across different continents, Jones et al. (2009) realized that subdivision Gp1 was responsible for almost 7.5% of all bacterial sequences. These results indicate that Acidobacteria is relatively abundant in sediments, including semi-arid mangrove sediments, although more studies are required for understanding the role of specific subgroups detected in mangroves. Bacteroidetes are found in numerous environments ranging from animal guts and skins to sediments and seawater (Thomas et al., 2011), always under anaerobic conditions. The ability of Bacteroidetes to occupy a variety of environments is mainly because they are recognized specialists in the degradation of high molecular weight organic matter, from polysaccharides to proteins. Some studies correlate the presence of Bacteroidetes with substrate pH, being significantly higher in alkali medium (Lauber et al., 2009). Although there was no significant difference in the pH of the studied mangroves, the relative Walderly Melgaço Bezerra, Tallita Cruz Lopes Tavares, Vanessa Lúcia Rodrigues Nogueira, Leonardo Ribeiro Oliveira Normando, Tatiana A. Bomfim, Alysson Lira Angelim, Vania Maria Maciel Melo abundance of this group was higher (11.4%) in MAL (pH 7.8) and lower (3.4%) in TIM (pH 7.0), confirming their preference for alkali habitats.
Few studies report an expressive presence of Chloroflexi in mangrove sediments (Liang et al., 2007;Pureza et al., 2013;Zhang et al., 2015), although in this study it represents 9% of the detected phyla. It was among the four phyla that were detected in all mangroves, being the second most abundant phylum in ICA (24%). In mangrove sediments from Shenzhen Bay, China, the relative abundance of Chloroflexi reached 28%, whereas Proteobacteria was the second most abundant phylum (Zhang et al., 2015). In mangroves from Southeastern Brazil, the relative abundance of that phylum varied from 5.4% to 1.3%, with the lowest value detected in contaminated mangroves (Andreote et al., 2012). Similarly, in this study, CEA mangrove showed the lowest relative abundance of green non-sulfur bacteria (Chloroflexi phylum), confirming this trend.

Diversity of bacterial genera
A total of 188 bacterial genera were identified among the 663 sequences from our dataset. So far, only 26 of these genera had been reported in mangroves (Baba et al., 2011;Basak et al., 2014;Chakraborty et al., 2015;Fan et al., 2012;Flores-Mireles;Winans & Holguin, 2007;Gao;Xu & Ruan, 2014;Gomes et al., 2010;Guo et al., 2011;Khatri et al., 2012;Mishra et al., 2012;Santos et al., 2011;Soto-Ramírez et al., 2008;Zhang et al., 2008;Zhao & Ruan, 2011). A great amount of randomly selected clones of this study was affiliated with groups that encompassed marine and/or estuarine anaerobic bacteria. The large inputs of organic matter support high rates of heterotrophic metabolism and, since oxygen is usually depleted below a few millimeters in mangrove sediments, even where the sediment surface is exposed to air, anaerobic metabolism predominates with decomposition mediated primarily by fermentative and sulfate-reducing bacteria (Alongi, 2010).
At least 33 genera were related to sulfur metabolism, comprising 25 genera of sulfatereducing bacteria (SRB) and eight genera of sulfur-oxidizing bacteria (Figure 3). Among the sulfur-oxidizing bacteria, the majority belongs to purple sulfur bacteria (Chromatiales), a group of anaerobic microorganisms that obtain energy from anoxygenic photosynthesis, i.e., oxidizing hydrogen sulfide to sulfite and sulfate. Ectothiorhodosinus was the most abundant genus of purple sulfur bacteria. Only one genus (Ignavibacterium) was assigned to green sulfur bacteria.
The iron cycle in mangroves is dynamic and closely linked with the sulfur cycle (Alongi, 2010). Iron and sulfur are important elements in the biogeochemistry of estuarine sediments because microbial groups related to both cycles play important roles in the decomposition of organic matter (Nóbrega et al., 2013). Sequences affiliated to Acidobacteria and related to iron metabolism were detected (Figure 3). Aciditerrimonas and Geothrix BACTERIOME ASSOCIATED WITH Rhizophora mangle SEDIMENTS WITHIN BRAZIL SEMI-ARID MANGROVES reduce ferric iron (III), whereas Acidimicrobium and Ferrimicrobium oxidize ferrous iron (II). Rhodobacteraceae, a highly diverse family within the class Alphaproteobacteria, known as a phototrophic iron oxidizer (Hedrich et al., 2011), was also well represented in the studied mangroves.
Besides the above-mentioned bacterial genera, this study reports the presence of wellknown genera related to the degradation of aromatic compounds, petroleum, and several contaminants (Figure 4). The genera Nitratireductor, Oceanicola, and Novosphingobium, associated with pyrene and naphthalene degradation (Lai et al., 2011;Suzuki & Hiraishi, 2007;Yuan et al., 2009), were detected in ICA, MAL, CEA, and ARA. Furthermore, bacteria belonging to the genera Anaeromyxobacter, Holophaga, Parvibaculum, Sphingopyxis, and Syntrophorhabdus, all related to degradation of phenolic compounds (Godoy et al., 2003;Liesack et al., 1994;Qiu et al., 2008;Cole & Tiedje, 2002;Schleheck et al., 2004), were detected in six of the eight studied mangroves (ACA, COR, MAL, COC, CEA, and ICA). The species Anaeromyxobacter dehalogenans, which is already extensively used in the bioremediation of uranium Sanford & Löffler, 2006), was detected in ACA and COR. The genera Devosia and Thalassolituus, encompassing bacteria related to the oil (Choi & Cho, 2013), hexane, and diesel (Ryu et al., 2008;Verma et al., 2009) degradation, were detected in COC and ACA. OTUs affiliated to Dongia, Hyphomicrobium (Alphaproteobacteria), Desulfatirhabdium, Desulfuromonas (Deltaproteobacteria), Nantrocella, Planococcus, and Steroidobacter (Gammaproteobacteria), related to the degradation of other hydrocarbons such as dyes, acetonitrile, dichloromethane, and steroid hormones (Balk et al., 2008;Engelhardt et al., 2001;Fahrbach et al., 2008;Liu et al., 2010;Sung et al., 2003), were found in all studied mangroves. The genus Haliea, previously reported by Santos et al. (2011) as a possible bioindicator for oil contamination in mangroves, was found in MAL, COC, CEA, ACA, and TIM, being more abundant in ACA (Figure 4). Altererythrobacter and Amorphous (Alphaproteobacteria), two genera of marine bacteria, were found in ICA, CEA, ARA, and ACA ( Figure 4). Some species belonging to these genera have been reported as producers of useful Figure 3 -Bacterial genera related to sulfur, iron, and nitrogen cycles detected by 16S rDNA cloning library of sediments collected in Rhizophora mangle root zones in mangroves (ICA, MAL, COC, ARA, ACA, COR, and TIM) from Northeast, Brazil enzymes and compounds such as epoxide hydrolase (Kumar et al., 2007), astaxanthin (Matsumoto et al., 2011), andexopolysaccharides (Wawonng et al., 2010). These results highlight the genetic potential of the bacterial community for biotechnological applications. Undoubtedly, researches undertaken in Brazil are at the forefront of studies on bioremediation and restoration of mangroves (Angelim et al., 2013;Santos et al., 2011;Sodré et al., 2013;Machado et al., 2019). However, knowing the mangrove microbiomes is essential to take further steps in this direction. The taxonomic characterization of those microbiomes allows us to (1) understand how environmental modifications related to specific anthropogenic impacts or related to climate change can alter those communities as well as make conclusions regarding their resistance and resilience, and (2) explore this bacterial composition in a strategic way to provide biotechnological tools to bioremediate and restore altered ecosystems (Allard et al., 2020).

CONCLUSIONS
The successive oil spills that occurred on the Brazilian coast in 2019 and 2020 illustrate quite well how important it is to know our biodiversity. Taking into consideration the bacterial diversity, this study provided a baseline to evaluate how it affected those coastal ecosystems, as well as provided tools that are being studied for the bioremediation of damaged mangroves.
Our dataset provides an important overview of the sediment bacteriome throughout the coastal zone in the semi-arid region of Brazil, comprising preserved and anthropized BACTERIOME ASSOCIATED WITH Rhizophora mangle SEDIMENTS WITHIN BRAZIL SEMI-ARID MANGROVES mangroves. The results endorsed a general pattern, a core of bacteria phylum-groups in mangroves, regardless of the local environment characteristics, and high redundancy of bacterial genera involved in the sulfur metabolism, especially sulfate-reducing bacteria. These organisms proliferate in habitats rich in partially oxidized substrates and anaerobic conditions, typically found inside red mangrove habitats. Furthermore, our study presents the identification of a signature of bacterial genera inside red mangrove habitats in the Northeastern Brazilian mangroves.